Many commercially important compositions, including fuel, lubricating oil, detergent, and plasticizer compositions, contain or are produced using oxygenate such as long chain alcohol (unbranched C4+ alcohol having one terminal hydroxyl). Butanol, for example, can be used to produce butyl acrylate and methacrylate. C5 alcohol can be used to produce lubricating oil additives such as zinc diamyldithiophosphate. Alcohols in the C6-C9 range can be used to produce plasticizer. C6-C9 alcohol, particularly C12-C15 alcohol, can be used to produce detergent, typically by first converting the alcohol to alcohol sulfate, ethoxylate, alcohol ether sulfate, or fatty amine.
One way to produce long chain alcohol is by catalytic hydroformylation of olefin to produce a reaction product comprising aldehyde, separating linear aldehyde from the reaction product, and then hydrogenation of the linear aldehyde. The hydroformylation reaction is carried out by contacting the olefin with a mixture of carbon monoxide and molecular hydrogen in the presence of at least one hydroformylation catalyst. Conventional hydroformylation catalysts typically comprise at least one metal selected from Groups 8-10 of the Periodic Table, especially at least one metal selected from Group 9, such as one or more of cobalt, rhodium, and iridium. Although rhodium and iridium exhibit greater hydroformylation activity, cobalt is typically used to lessen costs. When using a cobalt catalyst such as HCo(CO)4, hydroformylation conditions typically include a temperature ≥140° C. and a pressure ≥24 MPa. Relatively high pressure is needed for at least two reasons. First, high total pressure is needed to stabilize the catalyst at the reaction temperature. Second, selectivity for the desired linear aldehyde over branched aldehyde by-product increases with increasing carbon monoxide partial pressure.
Besides the relatively high temperature and pressure needed, conventional hydroformylation processes are also sensitive to the type of olefin present in the feed. Linear olefin is approximately an order of magnitude more reactive for producing the desired aldehyde than is iso-olefin having the same number of carbon atoms. Consequently, conventional hydroformylation processes typically require concentrating linear olefin in the feed, e.g., by separating and conducting away any isoolefin. In addition to these difficulties, catalysts such as HCo(CO)4 have a normal boiling point similar to that of desirable linear aldehydes, which increases the complexity of product and catalyst recovery stages.
In order to more easily recover catalyst from the long-chain alcohol, processes have been developed to produce long chain alcohol by direct hydration of 1-alkene using acids, metal oxides, zeolites, or clays. One difficulty in operating these processes results from an aspect of Markovnikov's rule: a proton bonds to a hydrocarbon molecule at the molecule's least-substituted carbon atom. Accordingly, the hydration reaction protonates the 1-alkene's double bond, which disfavors transition states amenable to long chain alcohol formation. Although triple-relay, platinum-ruthenium catalyst systems having anti-Markovnikov behavior have been developed for 1-alkene hydration, a more recent approach involves the direct production of long chain alcohol by a modified Fischer-Tropsch (“FT”) synthesis.
As reported in Lu, et al., ChemCat Chem 2014, 6, 473-478, processes using a templated Cu—Fe FT synthesis catalysts can be operated to strongly favor producing long chain alcohol (anti-Markovnikov behavior) from a carbon monoxide—molecular hydrogen mixture over producing branched alcohol having the same number of carbon atoms (Markovnikov behavior). The reference discloses reacting a 1:1 molar ratio carbon monoxide—molecular hydrogen mixture in the presence of a three-dimensional, ordered macroporous catalyst comprising CuO and Fe3O4. The reaction is carried out at a temperature of 240° C. and a pressure of 700 psi (approximately 4.8 Mpa). The reference discloses a selectivity to C2+ 1-alcohol production of 40% (weight basis) and a feed carbon monoxide conversion of 45% (weight basis). 1-alcohols in the reaction product have a number of carbon atoms in the range of from 1 to 16, with a weight average of about 9, resulting in an Anderson-Shulz-Flory Chain Growth Probability (“α”) of 0.74.
Catalysts are now desired for processes which produce oxygenate such as long chain alcohol from a carbon monoxide—molecular hydrogen mixture, the process having a feed carbon monoxide conversion >45% (weight basis) and a selectivity to oxygenate such as 1-alcohol >40% (weight basis). Catalysts are particularly desired which are effective for catalytically converting syngas to oxygenate at a reaction temperature ≤250° C. and a total pressure ≤5 MPa.